Imaging Hales Discontinuity beneath India: Seismological and Petrological Model

Hales discontinuity (H-D) in the Indian lithosphere had been previously imaged using P-wave receiver functions (P-RF) beneath some stations at a depth of ~75 km, one station (HYB) at a depth of ~90 km, and several stations without any observed H-D. No comprehensive explanation has been provided for these scattered observations. In order to reconcile such observations and to explore its genesis, we perform P-RF imaging and joint inversion with surface wave data to obtain well constrained depth and S-wave velocity ( Vs ) change beneath three stations in the Indian Shield. We use P-RFs from station HYB, calculated at Gaussian width (Gw) 2.5 (f max 0.46 Hz), to identify the H-D as the P-to-S converted phase (P h s) at ~11 s, with a preceeding phase at ~9.5 s, with opposite moveouts. Using crustal model for HYB we calculate arrival times of all crustal phases and show the ~9.5 s phase as Moho PpPp multiple. Our joint inverted mean model shows the H-D at 107±5 km depth beneath HYB with a ~3% Vs increase. Joint inversion of the low frequency (Gw 1.0, f max 0.18 Hz) P-RF stack, like previous studies, underestimates the depth and velocity contrast of H-D. For station GBA, we show that the 8-9 s phase, imaged earlier as the H-D, is a mid-crustal multiple , distinct from P h s on P-RFs computed for Gw 2.5. We observe P h s at ~10 s and model the H-D to be at 97±2 km depth with ~5% Vs increase. For station KOD, the P h s coincides with the negative mid-crustal reverberation for most ray parameters, causing destructive intereference. This explains the absence of P h s at several stations beneath India. We isolated a few ray parameter bins where P h s is distinct at ~10.5 s and modeled it at 101±5 km depth with 3% Vs increase. The most cited explanation for H-D, the spinel-garnet mineral transformation, cannot explain the observed ~3% increase in Vs . In order to explain the Vs jump, we calculated Vs of two types of mantle xenoliths (peridotite and eclogite), reported from the studied region, from their published bulk rock compositions using Perple_X. At the observed H-D depth and corresponding T defined by the Indian shield geotherm, the two rock types show a Vs difference of ~3%. Thus the H-D is interpreted as the discontinuity surface between upper mantle peridotite and paleo-subducted eclogitic oceanic slab.